The value of structural neuroimaging in the selection of patients for epileptic surgery

The aim of this study was to investigate the value of structural neuroimaging with MRI in the selection of patients for epilepsy surgery. We sought to determine whether MRI influenced decision concerning resective surgery and whether MRI provided much more useful information than enhanced CT.

Methods: Neuroimaging studies, MRI and CT, of 300 patients; 265 with partial and 35 with primary generalized seizures, evaluated for surgical treatment of epilepsy were analysed. The MRIs and CTs were interpreted using visual diagnostic criteria and findings were correlated with the EEG changes and clinical semiology.

Results: MRIs identified structural lesions in 142, CTs in 96 of all patients. The clinical semiology (partial seizures), MRI, CT and EEG focal findings were concordant in 72 cases. The group of 34 patients had resective surgery. The 7 patients were also operated with MRI and CT focal abnormalities discordant with EEG changes. Also one patient with primary generalized epilepsy and temporal lobe lesion (glioma) had resective surgery. MRI studies revealed structural lesions in 48 patients with normal CT studies. The 43 patients with partial epilepsy had normal CTs and lesions in MRIs; the 34 cases revealed correlation with the EEG findings in 29 temporal and 5 extratemporal regions. Surgery were performed in 23 cases. Also one with partial seizures and MRI detected hippocampal atrophy was operated, despite of generalized EEG patterns. In contrast CT revealed two patients with normal MRI and focal changes. The patients with partial seizures and only CT abnormalities (focal calcifications) were not operated due to discordant EEG findings. In group of 132 patients with normal neuroimaging studies and EEG identified seizure focus only 27 had anterior temporal lobectomy.

Conclusions: MRI studies gave additional information in case of 16% patients with intractable epilepsy in comparison with CT findings. Resective epilepsy surgery was almost twice as often performed when MRIs revealed structural abnormality. In operated patients, diagnostic sensitivity of structural MRI, CT and EEG to neurophatology were 70.6%, 46.7 and 92.4% respectively.